I was watching the CHM David Liddle interview, and I was surprised by his description of the “spark pen”, a 01970s pointing device using a glass panel and microphones, using the audio transit time of the sound of a spark to measure the position you were pointing at.
You can get several MHz of acoustic bandwidth through a glass panel, and a spark gap has submicrosecond rise time, so you can get submicrosecond positioning precision — maybe a millimeter or so. Two things occurred to me about this:
A very low-tech way to get the spark is by mechanically opening a switch with the pen point, pushing two pieces of metal apart, while running current through those pieces of metal in series with an inductor, ideally with regulated current. This is maybe easiest if the pen point doing the pushing is a glass rod, but metal would work too.
If you instead have a spark gap that can be induced to spark frequently by opening a high-voltage MOSFET in parallel with it, you can send a pseudorandom sequence of sparks that sounds like white noise, allowing you to track the pen’s position continuously instead of only on command.